5 ms·

Moore's Law was not a law, only a good fit of the data for a very long time. The ability of machines to replace people is advancing to points we've never seen

Moore's Law was not a law, only a good fit of the data for a very long time. The ability of machines to replace people is advancing to points we've never seen before, and I don't think any of us can really predict what will happen.

notahacker 9y ago
The ability of machines to replace people advanced to points we've never seen before in every other era either. And we got rid of whole other classes of jobs in the West for entirely different reasons like international trade (textile manufacturing) or culture (domestic servitude). What was persistently underestimated was people's willingness and ability to find new things which people wanted to pay for, many of them hitherto impossible without the ability to manufacture stuff at low cost, and many of them in areas which machines are still very bad at. Not being able to predict the future is a very strong argument against redesigning the welfare state to give more people money on the assumption that historic links between economic growth and the labour force are about to disappear.
- VikingCoder 9y ago
  Sound logic. Except you're presuming that Moore's Law will last forever. How will you know when it's about to stop?
  - notahacker 9y ago
    The people assuming Moore's Law will last forever are the people who are making "today I made some radiologists redundant, tomorrow not many jobs in medicine" connections I'm arguing "as things get cheaper because fewer people were used to make it, you'd better have really good reasons to believe people won't just use the excess money to buy more stuff as we've always observed them do in the past". If there's one resource that seems to consistently fulfil Jevons' Paradox it's labour. We might have fewer radiologists instead of more X-ray scans in future, but I see little reason to believe that won't be offset by the need for highly trained people to calmly reassure them about other diagnostic tools that have only just been invented. Throw in a bit of Veblen "people with excess money actually really like paying people to do things for them in unnecessarily laborious ways to signal how wealthy they are". It's rooted in human psychology as well as a couple of hundred years of empirical observations. I rather think the burden of proof - especially when calling for drastic action - is on the people arguing this time it'll be different, and they're the ones reliant on projections of trend lines to try to argue that computers that do everything better than humans are [still] only a decade away.
    - VikingCoder 9y ago
      "as things get cheaper because fewer people were used to make it" You're saying price drops when cost drops, and that's empirically not true. Price drops when demand drops, or when supply increases. When the cost drops, then the producer just gets more profit. And if they're wise, they buy or merge with their slightly less efficient competitors, because collusion is illegal, but no one is stopping M&A these days. And for necessary goods like food, water, shelter, clothing, it's not like the demand drops much over time, so the demand stays where it was (or grows with population.) Worker productivity is through the roof. That's automation. Real wages are going down. That's evidence. What happens when they keep dropping? "Nothing changes," is your answer?
      - notahacker 9y ago
        As I said, it's the neo-Luddites saying "but if you project these country-specific short term trend lines forward indefinitely, we'll all be unemployed...". Whereas we've got 200 years of data confirming the basic logic that the relative prices of basic food and clothing get much cheaper as lower costs increase competitors' ability to supply, something I'm genuinely amazed you can argue doesn't happen. Sure, individual companies can prop up their monopoly profit margins through IP protection or M&A or positioning themselves as a luxury item, but only in the same sense that water can be forced, in certain circumstances, to flow uphill.
        VikingCoder 9y ago
        Stop putting words in my mouth, and we can maybe have a real conversation. I'm not advocating protectionism, so that's not a fruitful branch of this conversation. Food and clothing are cheaper - that's great. The barriers to entry there are pretty low. How much cheaper is healthcare? Housing? Cars? College? Is there not enough competition in any of those? (Obviously not.)

notahacker 9y ago

The ability of machines to replace people advanced to points we've never seen before in every other era either. And we got rid of whole other classes of jobs in the West for entirely different reasons like international trade (textile manufacturing) or culture (domestic servitude). What was persistently underestimated was people's willingness and ability to find new things which people wanted to pay for, many of them hitherto impossible without the ability to manufacture stuff at low cost, and many of them in areas which machines are still very bad at. Not being able to predict the future is a very strong argument against redesigning the welfare state to give more people money on the assumption that historic links between economic growth and the labour force are about to disappear.

VikingCoder 9y ago
Sound logic. Except you're presuming that Moore's Law will last forever. How will you know when it's about to stop?
- notahacker 9y ago
  The people assuming Moore's Law will last forever are the people who are making "today I made some radiologists redundant, tomorrow not many jobs in medicine" connections I'm arguing "as things get cheaper because fewer people were used to make it, you'd better have really good reasons to believe people won't just use the excess money to buy more stuff as we've always observed them do in the past". If there's one resource that seems to consistently fulfil Jevons' Paradox it's labour. We might have fewer radiologists instead of more X-ray scans in future, but I see little reason to believe that won't be offset by the need for highly trained people to calmly reassure them about other diagnostic tools that have only just been invented. Throw in a bit of Veblen "people with excess money actually really like paying people to do things for them in unnecessarily laborious ways to signal how wealthy they are". It's rooted in human psychology as well as a couple of hundred years of empirical observations. I rather think the burden of proof - especially when calling for drastic action - is on the people arguing this time it'll be different, and they're the ones reliant on projections of trend lines to try to argue that computers that do everything better than humans are [still] only a decade away.
  - VikingCoder 9y ago
    "as things get cheaper because fewer people were used to make it" You're saying price drops when cost drops, and that's empirically not true. Price drops when demand drops, or when supply increases. When the cost drops, then the producer just gets more profit. And if they're wise, they buy or merge with their slightly less efficient competitors, because collusion is illegal, but no one is stopping M&A these days. And for necessary goods like food, water, shelter, clothing, it's not like the demand drops much over time, so the demand stays where it was (or grows with population.) Worker productivity is through the roof. That's automation. Real wages are going down. That's evidence. What happens when they keep dropping? "Nothing changes," is your answer?
    - notahacker 9y ago
      As I said, it's the neo-Luddites saying "but if you project these country-specific short term trend lines forward indefinitely, we'll all be unemployed...". Whereas we've got 200 years of data confirming the basic logic that the relative prices of basic food and clothing get much cheaper as lower costs increase competitors' ability to supply, something I'm genuinely amazed you can argue doesn't happen. Sure, individual companies can prop up their monopoly profit margins through IP protection or M&A or positioning themselves as a luxury item, but only in the same sense that water can be forced, in certain circumstances, to flow uphill.
      - VikingCoder 9y ago
        Stop putting words in my mouth, and we can maybe have a real conversation. I'm not advocating protectionism, so that's not a fruitful branch of this conversation. Food and clothing are cheaper - that's great. The barriers to entry there are pretty low. How much cheaper is healthcare? Housing? Cars? College? Is there not enough competition in any of those? (Obviously not.)